Host Specificity, Settling, and Metamorphosis of theTwo-tentacled Hydroid Proboscidactyla flavicirrata

RICHARD D. CAMPBELLl

THE COLONIAL HYDROID Proboscidactyla hasbeen found only on the leathery tubes of marinesabellid worms (Uch ida and Okuda, 1941:433 ;Hand, 1954; Brinckmann and Vannucci, 1965:367) . N othing is known about the means bywhich this specific commensal association arises.In this paper I present observations on planulasettling and metamorphosis, which indicate thatlarvae are caught in the tentacles of the sabellidworm and transferred to the rim of its tube.

METHODS AND MATERIALS

Medusae of P. flavicirrata (Brandt) weredredged in East Sound, Orcas Island, Washing­ton at a depth of 15 meters, on August 16-18,1964. Most specimens contained ripe gametes.Medusae kept in glass dishes without feedingshed and fertilized many eggs for three or fourdays at about 5 AM. Developing larvae were keptin sea water at 17°C, changed every 24 hours.

To test the influence of substrate on settling,five glass dishes of sea water were prepared withthe following: (1 ) 3 sabellid worm tubes fromwhich the worms had been removed; (2) 3tubes with worms; ( 3) 3 worms in glass tubes;(4) some perisarc of obelia, and (5) sea wateronly. About 50 one-week-old planulae werepipetted into each dish. Observations were madeduring the following 8 hours, and at intervalsover the next 8 days.

Studies on metamorph osis were made onplanulae which had settled on a tube (see be­low) , and which were transferred to a micro­scope slide where they completed metamor­phosis.

OBSERVATIONS

Planula settling and metamorphosis tookplace only on tubes containing sabellid worms,

1 Friday Harbor Laboratories, University of Wash­ington . Present address : Department of OrganismicBiology, Uni versity of California at Irvine, Irvin e,Californi a 92664 . Manuscript received May 11, 1967 .

regardless of whether the tubes were natural orartificial. N o settling occurred on tubes withoutworms, or on other surfaces.

Several hundred planulae were kept in clearglass dishes for 18 days. Du ring this periodnone metamorphosed or settled.

Observations on the behavior of the planulaeand sabellid worms during the settling processindicated the role of the worm. Initiation ofthe settling process began when a planula wascaught in the ciliary currents of the sabellid'sradioles (tentacles) . In the vicinity of theradioles these currents are much swifter thanthe planula's swimming movement; therefore,the planula must be considered as a passiveparticipant in initiating this association. How­ever, when the planula does contact the worm,nematocysts discharge and anchor the planula(Fig. 1) . The physical attachment is clearlyindicated when a single planu la binds to severaladjacent pinnules, clumping them.

The next stage in the settling process involvesthe transfer of the planulae from the radioles tothe rim of the tube. Th is is mediated throughretractions of the worm into its tube, which

FIG. 1. Proboscidatyla flavhirrata. Planula (ar­row) attached to two pinnules (p) of a sabeIIidworm. Pr inted from 16 mm film. Scale: 0.5 mm.

336

Proboscidactyla flavicirrata-CAMPBELL

scrape the planulae off onto the rim. This trans­fer was also made possible by a stickiness of theplanulae, which developed after the contactwith the worm. At least 20 minutes elapsed be­tween attachment to the pinnules and transferto the tube rim, and most retractions of theworm were unsuccessful in achieving thistransfer. Probably the elapsed time representsthe time required for secretion of an adhesivematerial.

Accidental contacts between the swimmingplanulae and the worm tube were refractile, andhad no effect on the activity of the planulae.However, after a planula had been transferrednormally to the tube it would underg o metamor­phosis even if removed from the tube.

The visible onset of metamorphosis occursabout 6 hours after settling . Figure 2 shows themature planula before attachment. Nematocystsare more abundant at the anterior end. Theplanula is characteristically spindle shaped.Figure 3 shows a planula several hours afterattachment, as the first visible signs of metamor­phosis become apparent. There is a loss ofrefractile quality of the endoderm at the futureoral pole. This reflects the formation of a welldelineated high columnar epithelium which ischaracteristic of the hypostome of the adultpolyp. Also at this time the mouth has begunto form. The animal in Figure 4 shows a pro­trusion in the body wall below the hypostomalregion. This evagination becomes a single ten­tacle, which elongates rapidly (Fig. 5) .

W ithin several hours another protrusion isseen developing on the body wall (Fig. 5) . Italways develops after the tentacle has been initi­ated, and usually after the tentacle has elongated.It is always on the same side of the body as thetentacle, and is generally situated in the middleof the body. This protuberance develops intothe "foot" (Campbell, 1967) which may behomologous with the stolon tip in other hy­droids. The segment of polyp posterior to thefoot adheres to the substratum, secretes a veryfine perisare, and becomes the stolon. The footmarks the anterior extent of the attachment tothe substratum, and the distal portion of thepolyp remains erect from the substratum.

During the next 24 hours, the observed,single-tentacled polyps began to glide along thesubstratum. The polyp column posterior to the

337

"foot" elongates during polyp movement, be­coming a stolon. Figure 6 shows a polyp justbeginning its movement. The stolon elongatesnot by terminal extension, for the aboral endremains fixed with respect to the substratum (orit may actually move in the direction of thepolyp) . Elongation is apparently due to stretch­ing by the advancing polyp. The mechanism ofthis advancement was not determined, but themovement appeared similar or identical to thatof mature polyps (Campbell, 1967) . It was al­ways oriented in the direction of the tentacleand foot.

About 30 young polyps were raised for morethan a week during which time they were fedsabellid eggs. N one of them during this timedeveloped a second tentacle. Of about 250metamorphosing polyps studied, however, 2possessed two tentacles just after metamor­phosis (Fig. 7) . The formation of these pairedtentacles was not observed.

After metamorphosis, the young polyps onthe worm tube were not oriented with respectto the axis of the tube, although the majoritywere right at the edge of the tube. Movementcarried some further from the rim. The behaviorof these young polyps was not observed for alonger time, and so it is not known whetherall of them were capable of forming colonies.

DISCUSSION

Interaction between the worm and the planulaappears to be a prerequisite to settling inProboscidactyla flavicil'l'ata. Since planulae de­prived of contact with worms did not undergometamorphosis during the more than 2 weeks ofobservation, the contact itself is probably astimulus for metamorphosis. It is possible thatnematocyst discharge is a direct part of thisstimulation. These conclusions explain how theclose association between the hydroid colony andworm tube, and the polyp's initial position onthe rim of the tube, may arise. It would be in­teresting to know if other species of worms orother animals, which must also draw Probosci­dactyla planulae into contact with themselves,similarly stimulate the planulae to settle andundergo metamorphosis. If there were a speciesspecificity involved, this intricate interactioncould present an explanation of the specificity

338 PACIFIC SCIENCE , Vol. XXII, July 1968

FIGS. 2-7. Proboscidactyla flavicirrata. FIG.. 2. Mature planula; scale: 0.1 mm. FIG. 3. Planula meta­morphos is, 3 hours after contact with sabellid worm ; oral pole is to the right ; the hypostomal endoderm (h)has differentiated ; 11, nematocyst ; scale: 0.1 mm. FIG. 4. Planula metamorphosis, 4.5 hou rs after contactwith worm ; t, tentacle rudim ent ; scale: 0 .1 mm. FIG. 5. Metamorphosis nearly complete, 7 hours after con­tact with worm ; t, rudiment of "foot" ; scale: 0.1 mm. FIG. 6. Beginn ing of migration ; i. "foot" ; tissueto the left of foot represents stolon (s), which in this case has stretched to about 3 times its original lengtb;12 hours after contact with worm ; scale: 0 .1 mm. FIG. 7. Ne wly metamorphosed polyp possessing two ten­tacles ; small protrusion at left is the "foot" ; scale: 0 .1 mm.

of the Proboscidactyla colony for the sabellidworm tubes.

The initial attraction of the planula to theworm site, however, appears to be quite non­specific, involving water currents set up by thebranchial cilia. In this respect the commensalspecificity resembles that of H ydractinia forshells inhabited by hermit crabs: Schijfsma

(1 935 :290-302) and Cazaux ( 1958 :2195) ,showed that there is no attraction of theHydractinia planula by the hermit crab, butrather that settling is apparently stimulated byparticular conditions of waterflow across a hardsubstratum, conditions which are frequentlypresented by a hermit crab shell in its habitatof swiftly moving water.

Proboscidactyla flavicirrata-CAM PB ELL

Proboscidnctyla metamorph osis resembles thatof other hydroids , despite the unusual finalmorph ology of the polyp. Metamorph osis occursonly after settling, which apparently providesa trigger for the development of polyp features.Settling occurs on the future aboral surface.Mouth , tentacles, and the hypostomal dome arethe first structures to appear, followed by theformat ion of the stolons more proximally onthe stalk.

A striking feature of the newly metamor­ph osed Ptoboscidactyla is its ability to moveacross the substratum. Adult polyps have thissame behavior pattern (Campbell , 1967), whichis essential to colony maintenance. The move­ments can be interpreted in terms of stolonelongation with the polyp residing near thestolon tip.

While in the mature Proboscidactyla colonymovement is directed distally along the wormtube, thus maintaining the polyp on the edgeof the tube as it is elongated , the movementof the newly metamorph osed larvae was notoriented with respect to the tube in the culturesobserved. Thus, in some instances this move­ment carried the polyp father from the tube rim.This may be unusu al behavior associated withlaboratory conditi ons, or it may be a normalevent which has an unappreciated selective ad­vantage.

The adult Proboscidactyla polyp has twotentacles. The newly metamorph osed larvae ob­served generally possessed only one. Althoughthe identity of the medusoid and polyp genera­tion of this species has not been established bycomplete life cycle studies in the laboratory(the most complete data are given by Uchidaand Okuda, 1941:435-439) , the identity ap­pears very pr obable. Presumably either theseyoung polyps were abnormal, or a second ten­tacle arises later. However, another possible ex­planation for the single tentacle is that onlythose polyps developing asexually, duringcolonial expansion, have two tentacles; single­tentacled polyps may generally start colonies.

339

SUMMARY

P!a.nulae of the hydroid Proboscidactylaflavtcll'rata are caught in ciliary currents ofsabellid worm radioles, attach to these radiolesby means of nematocysts, and are scraped offonto the rim of the leathery tube as the wormretracts. This explains how Proboscidactylacolonies may become initi ated on this site, andmay explain why the colonies are not foundelsewhere .

Metamorphosis, apparently triggered by in­teraction with the worm, is generally typical for~ydrozoans except that the final morph ologyIS unusual. A young polyp moves across thesubstratum, stretching the proximal end into atrailing stolon .

Most of the newly metamorph osed polyps ob­served had only one tentacle, although polyps innatural colonies normally have two.

REFEREN CES

BRINCKMANN, A., and M. VANNUCCI. 1965.On the life-cycle of Proboscidactyla ornata(Hydromedusae, Proboscidactylidae). Pubbl.Staz. Zool. N apoli 34 :357-365.

CAMPBELL, R. D. 1967. Colony growth patternin the commensal hydroid, Proboscidactylaflavicirrata. BioI. Bull. (In press.)

CAZAUX, C. 1958. Facteurs de la morphogenesechez un Hydraire polymorphe , H ydractiniaechinata Flem. C. R. Hebd. Seances Acad.Sci., Paris 247 ( 23) :219 5- 2197.

HAND, C. 1954. Three Pacific species of "Lar"(including a new species) , their hosts,medusae, and relation ships . (Coelenterate,Hydrozoa) . Pacific Sci. 8 (1) :51-67.

SCHIJFSMA, K. 1935. Observations on H)'­dractinia echinate (Flem.) and El/pagl/l'Usbernbardns (1.) . Arch. need. Zool.1(3) :261- 314.

UCHIDA, T. , and S. OKUDA, 1941. The hydroidLar and the medusa Proboscidactyla. J. Fac.Sci. H okkaido Univ. VI, 7(4 ) :431-440.